JAMES PRESCOTT JOULE 169
The solution of the zinc oxide in the sulphuric acid of the
cell does not contribute to the current, so he determines in
subsidiary experiments the quantity of heat evolved in this
reaction, and deducts it from the gross evolution of heat in
the cell. The corrections for losses by radiation and
conduction are also made.
The investigation of the heat production in batteries is
followed by an investigation of that in electrolytic cells.
He distinguishes between the resistance of conduction in
such cells and the resistance to electrolysis. He determines
the latter quantity first. After allowing for it, he again finds
that the heat evolution due to the resistance to conduction
obeys the same law.
He concludes that "if the electrodes of a galvanic pair of
given intensity be connected by any simply conducting body the
total voltaic heat generated by the entire circuit ((provided always
that no local action occurs in the pair) will., whatever may be
the resistance to conduction, be proportional to the number of
atoms (whether of water or of zinc] concerned in generating the
current"
He gives two more related conclusions, and then writes
that his results confirm Berzelius' suggestion "that the
light and heat produced by combustion are occasioned by
the discharge of electricity between the combustible and
the oxygen with which it is in the act of combination; and
I am of opinion that the heat arising from this and some
other chemical processes is the consequence of resistance
to electric conduction." His unpublished experiments on
the combustion of zinc turnings in oxygen, and Crawford's
explosions of mixtures of oxygen and hydrogen also support
the suggestion.
He is aware that Peltier has shown a current may pro-
duce cold. "I have little doubt, however, that the explana-
tion . . . will be ultimately found in actions of a secondary
character/'
His next four papers contain additions to his extra-
ordinary discovery of the quantitative relations between the
number of atoms, the heat and the electricity concerned in
the action of an electrolytic circuit